In the wake of the Stock Market Crash
of 1987, one of the most compelling questions for the nation's economy
was how the Federal Reserve might react. Would the Federal Open
Market Committee (FOMC)the policymaking arm of the Fedinject
more money into the economy? Was such a loosening of monetary policy
justified? Was a "wait-and-see" attitude more prudent?
Reliable economic statistics would not be available for weeks or
months; still, policymakers needed to determine what, indeed, was
going on with the economy in the days following the Crash. On what
information would the Fed base its decision making?
A November 1987 Business Week article provided
an answer: "Thousands of ... tidbits have poured into the
Federal Reserve System's Washington headquarters since Bloody
Monday. ... The regional Feds survey businesses in their districts,
tapping more than 300 members of various Fed boards, as well as
hundreds of informal contacts, to compile the 'Beige Book' on
regional business conditions. ... Even 'eyeball evidence'like
Minneapolis Fed President Gary H. Stern's car counts at the local
mallsgo into the information stream."
For many Americans, this reference was their introduction to
the Beige Book, a previously obscure document that had only been
public since 1983, but that would soon become a regular feature
of business news pages and would take on oracular properties for
many media and Fed watchers. Briefly, the Beige Book is a largely
anecdotal compilation of economic reports from each Federal Reserve
district, from which a national summary is then drawn, and which
is submitted to the Federal Reserve Board and released to the
public two weeks prior to FOMC meetings, or eight times a year.
The Beige Book is just one piece of information used in the making
of monetary policy; the FOMC relies most heavily on forecasts
generated from national models and on more current data and information
than the Beige Book contains. Still, the release of this document
prompts many to search for clues to Fed policy amid the mostly
matter-of-fact recitation of current economic conditions. But
is such faith in the prophetic powers of the Beige Book justified?
Also, how did such a document develop and how did it come to take
on such importance?
An analysis by researchers at the Federal Reserve Bank of Minneapolis
suggests an answer to the first question: While the gathering
of regional information for the Beige Book provides value to the
FOMC as a reflection of the economy, the national summary based
on the compilation of regional reports does not improve upon private
sector forecasts. Consequentlybecause the Beige Book does
not improve upon private sector forecasts, and because the FOMC
looks at an array of forecasts and national indicators, and Board
staff generates its own forecaststhe Beige Book is not a
good indicator of the future course of monetary policy. But before
we continue an investigation into the reliability and value of
the Beige Book as a predictor of economic growth, it is useful
to recount the history of this document and its use by the media.
[For those curious about the incident described above and who would
like to investigate the Fed's post-Crash Beige Books for themselvesas
well as periods dating back to 1970, when such records were first keptplease
visit the only electronic archive of Beige Books and
Red Books (as they were known prior to 1983). Current Beige Books are
also placed on the Web site upon release to the public.]
E Pluribus Unum
When a Federal Reserve bank president gives a speech about the
economic prospects of his particular district, he brings a wealth
of information to the podiumboth qualitative and quantitativegathered
by his staff of economists and researchers. This regional focus
reassures the president's listeners that he has his hand on the
pulse of the district's economy and that he will convey this message
to the FOMC, and also ensures that the FOMC has viewpoints representative
of the entire country. To varying degrees, it has always been
The FOMC was formed by the Banking Act of 1933 (and did not
include voting rights for the Board of Governors), changed in
the Banking Act of 1935 to include the Board of Governors and
to closely resemble the present-day FOMC, and amended in 1942
to give us the current voting structure, which consists of the
seven members of the Board of Governors, the president of the
New York Fed and four other Fed presidents who serve on a rotating
basis. These Acts were an attempt, among other things, to centralize
the Fed's policymaking while preserving input from Federal Reserve
bank presidents. (It should be noted that while Federal Reserve
bank presidents vote on a rotating basis, they all attend each
FOMC meeting and contribute to the debate on monetary policy.)
The early FOMC at first met quarterly to consider its business;
today, the FOMC meets eight times a year, but decisions regarding
monetary policy are not limited to formal meeting dates, as the
chairman can call a teleconference of the FOMC at any time.
This system for making monetary policyregional viewpoints
in the making of national policyis one of the hallmarks
of Fed structure and has not been without its critics over the
years. From the beginning, opinions differed on the need for-and
the location of-geographic representation on the Board, and the
debate continued with the formation of the FOMC. Also, debate
about Fed structure is never far from Capitol Hill; in 1964, for
example, hearings were held that considered, among other things,
abolition of the FOMC. In his testimony, then Board Governor George
Mitchell said: "I think that regional representation from
men whose day-to-day business activities keep them in touch with
industrial, commercial and banking developments in the major centers
of the Nation brings to the committee qualitative judgments and
insights that aggregative statistics will always lack."
Of course, the importance and dominance of national policy over regional
considerations is understood; that is, the FOMC would not alter monetary
policy to address an economic concern pertinent to just one district.
But, as Mitchell testified and many other FOMC members have suggested,
regional input still plays a role in the formulation of that policy. As
Fed Governor Laurence Meyer said in a 1998 speech: "The presidents,
in addition to having regional information, also tend to have real-time
information about consumer spending, business investment, and wage and
price developments, for example, gathered from speaking to firms in their
districts." (Reprinted in The Region, June 1998.)
For the purposes of the Beige Book, Federal Reserve banks receive
such information from a variety of sources: boards of directors,
advisory councils on regional business activity, branch bank directors
and a network of contacts from business, agriculture and labor.
Each bank then prepares a report on the economic conditions of
its district, and from the compilation of these reports a national
summary is distilled.
From Red to Beige and Private to Public
Prior to 1970 and the arrival of Arthur Burns as the chairman
of the Federal Reserve Board, the FOMC made comments in a set
pattern, known as a "go-around." Burns was not a fan of this formalized
process. Also, Burns was not a consensus builder when it came
to making monetary policy, as was his predecessor, William McChesney
Martin, who listened to everyone's input before making his decision.
"Burns did exactly the reverse. He began his first meeting
by saying he would not be bound by the old ways and that there
would be no set order for discussion," according to Donald
F. Kettl's Leadership at the Fed (Yale University Press, 1986).
Burns also decided it would be a more efficient use of the FOMC's
time to have the reports on district conditions prepared in advance
and compiled for the Committee's edification. Burns' directive
formalized and broadened the information-gathering process, and
thus was born the Red Book, which was the predecessor to the Beige
Book. The FOMC minutes of May 5, 1970, shed further light on Burns'
intent: "Chairman Burns observed that one purpose of the
new report, as he understood it, was to permit the Committee to
draw to a greater extent than at present on the knowledge of Reserve
Bank people, including the directors. The emphasis was placed
on qualitative information, such as opinions and judgments ..."
Aside from the color of their covers, the Red and Beige books
differed in one important way: The Red Book was prepared for policymakers
only, and was not intended for public consumption. Despite this,
readers of the Red and Beige books will not find much difference
in content, although during the early years there was more individuality
in the various reportsin terms of length and, occasionally,
The Red Book became public in 1983 following a request by the
long-time delegate to the U.S. House of Representatives for the
District of Columbia, Walter E. Fauntroy. Actually, what Fauntroy
requested was that the Green Bookwhich contains the Fed's
national models and economic forecastsbe made public. However,
the Board deemed this unwise and the Red Book was offered in its
place. To mark the change, the color red was dropped and beigeit
was for a time also called the Tan Bookadorned the new cover.
The text of the new document was little changed, although care
was taken to delete any references to specific businesses or names;
Board staff also began to review the Beige Book prior to its release
to the public and to Fauntroy's House subcommittee.
However, one other significant change was made: To detract from
the implied importance of the document in FOMC policymaking, the
release of the Beige Book was timed for two weeks prior to an
FOMC meeting. It was thought that by building in a two-week gap,
the media and others would recognize that the information was
not timely and, therefore, did not have a major influence on policy.
The Media Discovers the Beige Book
For a number of years, concern about a possible overinterpretation
of the Beige Book was moot because little attention was paid to
the document. But that changed following the Stock Market Crash
of 1987 and the subsequent attention that was focused on the Fed.
An electronic search of the Wall Street Journal, New York Times and Washington Post reveals
little or no mention of a Beige or Tan Book prior to 1987, after
which the references climb from six articles in 1988, to 18 five
years later and 45 in 1998. Even as late as December 1988, an
article in the Wall Street Journal suggests that
the public was still becoming accustomed to the report: "A
number of analysts mentioned the 'Tan Book,' a publication of
the Federal Reserve Board, as having a negative impact on both
gold and silver ..." (WSJ, Dec. 1, 1988.)
That reference also suggests that the public had already begun
reading the tea leaves of the Beige Book to make their moves in
the market. But some traders were savvy to the report, according
to another story in the same issue of the Wall Street Journal:
"Traders said the report eased fears among many investors
that the Fed would push interest rates higher to cool the economy.
But some market strategists suggested that complacency could be
costly. They said the report probably won't carry much clout with
the Fed decision makers, who are looking at more current data."
But this wasn't universally understood. Earlier in 1988, a March
piece in the Washington Post, which reads as a sort
of introduction to the Beige Book and Fed information gathering
in general, suggests that the report was influencing the markets:
"... stock traders and other financial market participants
decided the book's message was one of greater strength in the
economy than they had been expecting. To the markets, that meant
that the Fed was not likely to seek lower interest rates any time
soon, and the traders' actions sent long-term interest rates up
and bond prices down." (WP, March 27, 1988.)
The Beige Book's luster as a crystal ball hadn't dimmed by 1993,
according to a December Wall Street Journal story:
"Prices recovered in the afternoon following release of the
Federal Reserve's Beige Book report, which provides anecdotal
evidence of economic activity around the country. 'The Beige Book
had kept the market a little bit concerned through the morning,'
Mr. Alexy said. 'But it basically said economic activity is expanding
at a moderate pace and it cited prices as being very stable. That
was the kind of outlook the market was able to deal with.'"
(WSJ, Dec. 9, 1993.)
But is that the kind of outlook for which the Beige Book was
intended? More such citations are available for 1993, but let's
jump now to 1998 to see how the media and Fed watchers have used
the Beige Book in recent months. From a Dec. 10, 1998, Reuters
report in the New York Times: "The overall market
rallied after the Federal Reserve said in its beige book summary
of economic conditions that the nation's economic expansion continued
in November ..." Another article in that same issue says
that the Beige Book "could point to a future cut in interest
A November Wall Street Journal article suggests
that the recent Beige Book will likely be cited as one of "the
bits of evidence" at the upcoming FOMC meeting "by those
who want to cut short-term interest rates again." (WSJ, Nov.
In March 1998 the following report appeared in the Wall
Street Journal: "For most of the session, bond prices
were at a virtual standstill, hovering close to the previous day's
closing levels. However, a somewhat hawkish tenor in the Fed's
latest Beige Book, a periodic report on economic conditions across
the 12 Federal Reserve districts, pushed the market down slightly."
(WSJ, March 19, 1998.)
And so, roughly 15 years after it was made public, and nearly
30 years after it began as a document meant to mirror the condition
of the economy, the Beige Book has become a crystal ball used
to predict not only the direction of the economy, but also the
response of monetary policy.
The Minneapolis Fed Study
Of course, all of these citations (and many more left unmentioned
in other newspapers and in other media) beg the question: How
much attention should the media, Fed watchers and the general
public pay to the Beige Book as a predictor of the national economy
and monetary policy? The short answer isnot much. The longer
answer is provided by research at the Minneapolis Fed that investigated
two questions: First, does the Beige Book national summary help
predict growth in real output during the quarter in which the
Beige Book is released? Second, does the Beige Book national summary
improve upon the real growth predictions of professional private
Since both of these questions are quantitative, rather than qualitative,
we needed to convert the qualitative information in the Beige Book into
a quantitative measure. Two economists at the Minneapolis Fed assigned
consistent quantitative measures to each national summary since the Red
Book started in 1970.1 We
will refer to this quantitative measure as the Beige Book score.2
We first examined whether the Beige Book score provided any help in
predicting growth in real output during the quarter in which the Beige
Book was released. Regression analysis showed that the Beige Book score
was statistically significant in predicting current quarter real growth.
The Beige Book score, by itself, explained about 30 percent of the uncertainty
in current quarter real growth.3 (See chart.)
|Real GDP Growth: percentage computed
from initial estimates of real GDP.
||Beige Book Output Score:
For overall output, economists assigned a score between +2
and -2 in increments of 0.5 with +2 indicating very strong
and expanding economic conditions and -2 a very weak and declining
But this leads us to our second, and more important, question:
Given private sector forecasts, does the Beige Book really add
to our knowledge of the economy? Does it provide any new quantitative
information about current economic conditions that is not already
captured in the predictions of private sector forecasts made at
the same time?
Professional private sector forecasters have made predictions about
real growth, inflation and other economic variables for more than 30 years.
Recent research suggests that private sector forecasts are quite accurate.4 Since private sector forecasters look at much of the same economic data
that the Fed does in compiling the Beige Book, their forecasts may already
capture what we know about economic conditions at a particular time. Private
sector forecasters have a strong economic incentive to make accurate predictions,
and we should expect that they would seek out any relevant information
to help them make accurate forecasts.
Because we wanted to determine whether the Beige Book gives us more
information about economic conditions than private sector forecasts alone,
we needed data on private sector forecasts since the Red Book's inception
in 1970. The only quarterly survey of private sector forecasters that
began before the Red Book is the American Statistical Association-National
Bureau of Economic Research Survey of Professional Forecasters, which
began in 1968.5 We found
that the ASA-NBER prediction of current quarter real growth, by itself,
explained about two-thirds of the uncertainty in real growth. That is,
the private sector forecasts explained more than twice as much of the
uncertainty in current quarter real growth as the Beige Book score alone.
In answer to our second question, our analysis showed that the Beige
Book does not provide more accurate predictions of current quarter real
growth once the private sector forecasts have been taken into account.
We found that in a regression of current quarter real growth on the private
sector forecast and the Beige Book score, the coefficient on the Beige
Book score is not statistically significant.6 Therefore, while the Beige Book does have some predictive
power, the media and Fed watchers would do well to put aside the Beige
Book and focus on private sector forecasts in their attempts to predict
So What is the Value of the Beige Book?
As described above, the gathering and reporting of regional information
has evolved from a collection of anecdotal information presented
orally at FOMC meetings, to a more systematic compilation of regional
business and economic information. Still, even today, in an era
of national economic models and a wealth of economic data, this
formalized Beige Book process is largely anecdotal and is viewed
with a certain amount of skepticism by some academics. Alan Blinder,
former Fed vice chairman, once characterized this process as the
"ask your uncle" method of gathering information. (Journal
of Economic Perspectives, Spring 1997.)
However, there may be some merit in asking your uncle. Such
"reality checks" may give policymakers the "stories"
that help them make sense of the statistics and models that they
pore over prior to each FOMC meeting. This context can then become
a communicative tool for the Fed, so when policymakers explain
their actions to the public they can place their decisions within
the framework of the "real" economy and not just some
abstract models. Also, there are times when the most sophisticated
economic models can't bring insight to a current economic phenomenon.
In such an instance, asking your uncle may provide helpful insight.
Of course, the uncle in this case is more valuable than the derogatory
term implies: The Fed gathers its Beige Book information from
hundreds of sources across the country, andas noted at the
beginning of this articlethis process certainly proved instructive
in the days and weeks following the 1987 Stock Market Crash.
In the end, our analysis leads us to conclude that, in the case
of the Beige Book, the parts are greater than the whole. This
may seem contradictory. How can regional reports be of value when
the measure of those reports, the national summary, doesn't help
us make policy? Maybe it's worthwhile to turn that question around.
How can we make policy without knowing what is really occurring
in the economy? Also, to quantify and aggregate the information
in the Beige Book, which was never intended for quantification
and aggregation, may not do the Beige Book justice. The Beige
Book may tell stories that go beyond the numbers. The economist
Robert E. Lucas once wrote: "How does one acquire knowledge
about reality by working in one's office with pen and paper?"
(Econometrica, March 1993.) Pen and paper models
can tell us much about what is going on in the economy, but this
does not deny that it is beneficial to look out the window and
observe what is really happening.
David S. Dahl, Public Affairs economist, Edward Lotterman,
former Minneapolis Fed regional economist, and Stanley L. Graham,
economist and Minneapolis Fed adviser, analyzed the Beige Books and Red
Books used in the study described in this article; they also offered analysis
in the preparation of this report.
1 To quantify and evaluate Beige Book information, two Minneapolis Fed economists
read and scored the national summary of 265 Red/Beige books. Scores were
given for economic condition and change for six aspects of the economy-output,
labor markets, inflation, consumer spending, construction and manufacturing.
For economic conditions, each aspect was given a score between +2 and
-2 in increments of 0.5, with +2 indicating very strong and expanding
conditions, and -2 indicating very weak and declining economic conditions;
for other scores: 1 indicated that the rate of growth accelerated, 0 meant
no change and -1 that growth decelerated. With regard to the condition
score for output, which is a proxy for real gross domestic product, one
economist scored 263 out of 265 reports, another 241.
2 As noted
in this article, the Beige Book was a confidential document between 1970
and 1983 (when it was known as the Red Book); one of the questions under
investigation in this analysis was whether the predictive power varied
between the Red and Beige Book eras. The answer: There was no important
difference in the predictive power of the scored national summary between
the Red Book period and the Beige Book period.
1992, the official measure of real growth was real GNP growth. Since 1992,
the official measure of real growth has been real GDP growth. Our statistical
analysis compared the Beige Book score with the initial estimate of real
growth for the quarter. We chose the initial estimate that corresponded
to the official measure of real growth in each quarter.
and Runkle show, for example, that the price forecasts of professional
forecasters appear to be neither too high nor too low on average. (Michael
P. Keane and David E. Runkle, "Testing the Rationality of Price Forecasts:
New Evidence from Panel Data," American Economic Review, vol. 80, no.
4, September 1990, pages 714-735.) Those forecasts also appear to accurately
take into account the effect of past inflation, unemployment, oil price
changes and other economic variables on future inflation.
1990, the Federal Reserve Bank of Philadelphia has conducted this survey.
Another well-known survey, Blue Chip Economic Indicators, started releasing
quarterly forecasts of real growth in 1978, eight years after the Red
6 We should
note that this result differs somewhat from a recent Dallas Fed study.
(Nathan S. Balke and D'Ann Petersen, "How Well Does the Beige Book Reflect
Economic Activity? Evaluating Qualitative Information Quantitatively,"
Research Department Working Paper 98-02, Federal Reserve Bank of Dallas,
June 1998.) This study found that a Beige Book score was statistically
significant, but it only improved the forecast by a small amount. Moreover,
we think the Dallas Fed study overestimated the predictive power of the
Beige Book score because of a methodological error. The Dallas study assumes
that the error term was not correlated over time, but it was. This implies
that the Dallas regression model will overstate the true predictive power
of the Beige Book score.